49 6 
A rber and Parkin. — Studies on the 
Yet it is undoubted that, of all Gymnosperms, the Gnetales stand 
nearest to the Angiosperms. The number of characters in which they 
approach the Angiosperms is such as to render improbable, for example, 
Lotsy’s conclusion that they have no nearer common ancestor than the 
Pteridophytes. 
Several attempts have also been made, as we have seen, to link the 
Gnetales directly with certain existing Dicotyledonous families. None 
of these have as yet carried conviction, nor do we think it likely that 
these views will gain ground in the near future. 
The deductions drawn from the superficial resemblance of the aggre- 
gates of the fructifications, and certain vegetative features, to the amenti- 
ferous and some other apetalous Dicotyledons have in our opinion been 
pushed too far. The naked ovule, alone, is a bar to any direct linkage. 
It remains naked, despite all attempts to clothe it in a carpellary covering. 
The old suggestion of a close relationship between the Gnetales and 
the Coniferae, especially to Taxus , scarcely demands attention now. 
The Gnetales thus at the present time remain largely a phylogenetic 
puzzle. They are Gymnosperms, but, as we have seen, they undoubtedly 
possess strong Angiospermous affinities. Yet the attempts which have 
hitherto been made to derive the Angiosperms from them have not, it 
must be admitted, the merit of simplicity. 
Morphological Considerations. 
In the present consideration we shall confine our attention chiefly to 
the fructifications, for, just as in the case of the Angiosperms, we regard 
these as forming the crux of the whole problem. Conclusions in regard to 
anatomy, or even those derived from a study of the embryo-sac, however 
interesting, we look upon in this case as of less phylogenetic value. 
We propose to inquire whether the Strobilus Theory, as outlined in 
our earlier paper in reference 'to the Angiosperms, will not afford a satis- 
factory explanation of the Gnetales. We may, however, first recapitulate 
the new terminology introduced in that memoir, which we propose to 
apply here also. 
We restrict the use of the term ‘ flower ’ to the Angiosperms. The 
fructifications of the Gnetales, often referred to as ‘ flowers ’, we shall speak 
of as strobili . 1 * The inflorescences, which in the case of Welwitschia and 
Ephedra have been frequently termed cones, we shall refer to simply as 
aggregates of strobili. We prefer also the more non-committal terms, 
micro- and megasporophylls, in place of stamens and carpels, and amphi- 
sporangiate and monosporangiate for hermaphrodite and unisexual. 
1 We have already shown that the typical, and more primitive Angiospermous flowers are 
strobili [Arber and Parkin (’07), p. 36]. 
